AI-Berkshire: A Python Framework for Value Investing Using Multi-Agent AI
The Python-based repository xbtlin/ai-berkshire has gained significant popularity on GitHub, surpassing 13,000 stars. It introduces a value investing research framework that leverages Claude Code and multi-agent adversarial analysis to apply the investment methodologies of Warren Buffett, Charlie Munger, Duan Yongping, and Li Lu. This project demonstrates the growing intersection of generative AI and quantitative finance, showing how agentic workflows can automate complex fundamental analysis. By simulating adversarial debates between AI agents, it aims to reduce bias and improve the rigor of investment decision-making. The framework is built to work with Claude Code and implements parallel multi-agent research. It models the investment philosophies of four legendary value investors to evaluate financial data and debate investment theses.
## BACKGROUND
Value investing is an investment paradigm that involves buying securities that appear underpriced based on fundamental analysis. Claude Code is an agentic coding tool developed by Anthropic that operates in the terminal to understand codebases and execute tasks, while multi-agent systems use multiple AI models interacting with each other to solve complex problems.